Learning Generalized Relational Heuristic Networks for Model-Agnostic Planning

TL;DR

This paper introduces a neural network-based method for learning generalized heuristics for goal-directed planning that are transferable across different object types and quantities, improving efficiency and applicability.

Contribution

It presents a novel approach for learning heuristics without symbolic action models, using abstract state representations for better generalization across diverse planning problems.

Findings

Heuristics are transferable to larger problems with different objects.

The method outperforms prior approaches in generalization.

Empirical results on benchmark domains demonstrate effectiveness.

Abstract

Computing goal-directed behavior is essential to designing efficient AI systems. Due to the computational complexity of planning, current approaches rely primarily upon hand-coded symbolic action models and hand-coded heuristic-function generators for efficiency. Learned heuristics for such problems have been of limited utility as they are difficult to apply to problems with objects and object quantities that are significantly different from those in the training data. This paper develops a new approach for learning generalized heuristics in the absence of symbolic action models using deep neural networks that utilize an input predicate vocabulary but are agnostic to object names and quantities. It uses an abstract state representation to facilitate data efficient, generalizable learning. Empirical evaluation on a range of benchmark domains show that in contrast to prior approaches, generalized heuristics computed by this method can be transferred easily to problems with different objects and with object quantities much larger than those in the training data.